# cherry-studio v2.0.0-rc.4

- Product: cherry-studio (https://whatsnew.fyi/product/cherry-studio)
- Vendor: cherry-studio
- Date: 2026-08-02
- Version: v2.0.0-rc.4
- Original notes: https://github.com/CherryHQ/cherry-studio/releases/tag/v2.0.0-rc.4
- Permalink: https://whatsnew.fyi/product/cherry-studio/releases/v2.0.0-rc.4
- Labels: Pre-release

What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog.

Reuse: the summaries, labels and curation here are © What's New. Quote freely with attribution and a link back; wholesale republication of the corpus is not permitted — terms: https://whatsnew.fyi/terms. The vendors' own release notes remain their publishers'.

---

- **added** — Add bounded chunk reads for file IPC
- **added** — Allow changing the default endpoint in provider settings
- **added** — Add Radeon Cloud integration
- **added** — Show folder import copy progress in knowledge
- **added** — Restore inline web search and knowledge base citations in chat
- **added** — Support DeepSeek V4 Flash Responses API in provider registry
- **added** — Stream PDF previews via range reads
- **added** — Generate contextual empty-state greetings in chat
- **added** — Implement content-based dedup detection for internal files
- **added** — Implement policy-based file entry garbage collection with cleanup_policy and scan reaper
- **fixed** — Unify file and artifact preview routing
- **fixed** — Stop defaulting new agents to auto in permission mode
- **fixed** — Distinguish mini app launchpad icon from window pin
- **fixed** — Reset stale filesystem targets in agent migration
- **fixed** — Sync multi-model context selection in chat
- **fixed** — Keep the mode warning out of single-line rows in agent permission mode
- **fixed** — Handle missing Claude invocation models in agent runtime
- **fixed** — Match global model metadata for custom providers in data migration
- **fixed** — Open paintings on an empty page
- **fixed** — Reject invalid API key characters in provider settings

##### What's Changed
* feat(file-ipc): add bounded chunk reads by @eeee0717 in https://github.com/CherryHQ/cherry-studio/pull/17152
* fix(file-preview): unify file and artifact preview routing by @zhangjiadi225 in https://github.com/CherryHQ/cherry-studio/pull/17646
* refactor(design-tokens): converge semantic color contract by @zhangjiadi225 in https://github.com/CherryHQ/cherry-studio/pull/17535
* chore(renderer): remove dead v1 prompt tool-call residue by @vaayne in https://github.com/CherryHQ/cherry-studio/pull/17686
* test(renderer): complete remaining frontend test convergence by @zhangjiadi225 in https://github.com/CherryHQ/cherry-studio/pull/17676
* feat(provider-settings): allow changing the default endpoint by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/17691
* feat(provider): add Radeon Cloud integration by @JoursBleu in https://github.com/CherryHQ/cherry-studio/pull/16622
* fix(agent-permission-mode): stop defaulting new agents to auto by @vaayne in https://github.com/CherryHQ/cherry-studio/pull/17690
* feat(knowledge): show folder import copy progress by @eeee0717 in https://github.com/CherryHQ/cherry-studio/pull/17655
* fix: distinguish mini app launchpad icon from window pin by @Pleasurecruise in https://github.com/CherryHQ/cherry-studio/pull/17683
* feat(chat-citations): restore inline web search and knowledge base citations by @eeee0717 in https://github.com/CherryHQ/cherry-studio/pull/17590
* test(renderer): consolidate frontend regression coverage by @zhangjiadi225 in https://github.com/CherryHQ/cherry-studio/pull/17671
* fix(agent-migration): reset stale filesystem targets by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/17687
* fix(chat): sync multi-model context selection by @ousugo in https://github.com/CherryHQ/cherry-studio/pull/17371
* feat(provider-registry): support DeepSeek V4 Flash Responses API by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/17694
* fix(agent-permission-mode): keep the mode warning out of single-line rows by @vaayne in https://github.com/CherryHQ/cherry-studio/pull/17698
* feat(file-preview): stream PDF previews via range reads by @eeee0717 in https://github.com/CherryHQ/cherry-studio/pull/17153
* fix(agent-runtime): handle missing Claude invocation models by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/17670
* feat(chat-greeting): generate contextual empty-state greetings by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/17408
* fix(data-migration): match global model metadata for custom providers by @tommyzhang100504 in https://github.com/CherryHQ/cherry-studio/pull/17692
* fix(paintings): open on an empty page by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/17660
* fix(provider-settings): reject invalid API key characters by @Pleasurecruise in https://github.com/CherryHQ/cherry-studio/pull/17678
* feat(file-manager): content-based dedup detection for internal files by @EurFelux in https://github.com/CherryHQ/cherry-studio/pull/17210
* refactor(file-ipc): audit legacy file IPC and begin superseding FileStorage by @EurFelux in https://github.com/CherryHQ/cherry-studio/pull/16735
* feat(file-manager): policy-based file entry GC with cleanup_policy and scan reaper by @EurFelux in https://github.com/CherryHQ/cherry-studio/pull/16727
* test(job,stream-manager): remove wall-clock races from cancel and idle-timeout tests by @EurFelux in https://github.com/CherryHQ/cherry-studio/pull/17707
* fix(onboarding): skip redundant privacy agreement writes by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/17708
* fix(ui): keep focus feedback within control bounds by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/17436
* refactor(ui): polish settings, chat, and resource interfaces by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/17710
* fix(agent-messages): prevent virtual list blanks after send by @zhangjiadi225 in https://github.com/CherryHQ/

_[Truncated at 4000 characters — full notes: https://github.com/CherryHQ/cherry-studio/releases/tag/v2.0.0-rc.4]_